fix(studio): pin llama.cpp to b8637 (Gemma 4 support)#4796
fix(studio): pin llama.cpp to b8637 (Gemma 4 support)#4796danielhanchen merged 1 commit intomainfrom
Conversation
ggml-org/llama.cpp b8637 includes Gemma 4 support (ggml-org/llama.cpp#21309). Revert the temporary "master" default back to a pinned release tag. This eliminates the HTTP 422 errors from the prebuilt resolver (which could not find a release matching "master"), avoids unnecessary source builds, and restores prebuilt binary downloads on all platforms.
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request updates the default llama.cpp tag from master to b8637 across the Python, PowerShell, and Shell setup scripts. While this change addresses HTTP 422 errors associated with the master branch, feedback indicates that prebuilt resolution will still fail and fall back to source builds because the official repository lacks the required manifest file.
|
|
||
|
|
||
| DEFAULT_LLAMA_TAG = os.environ.get("UNSLOTH_LLAMA_TAG", "master") | ||
| DEFAULT_LLAMA_TAG = os.environ.get("UNSLOTH_LLAMA_TAG", "b8637") |
There was a problem hiding this comment.
While pinning to b8637 correctly avoids the HTTP 422 errors associated with the master ref, prebuilt resolution will likely still fail. The logic in resolve_install_release_plans requires a llama-prebuilt-manifest.json asset (line 1182), which is absent from official ggml-org/llama.cpp releases. Consequently, the script will continue falling back to source builds unless DEFAULT_PUBLISHED_REPO (line 65) is reverted to a repository containing the expected Unsloth metadata, or the resolution logic is updated to support manifest-less releases.
|
Updated to use The current latest release (b8637) includes Gemma 4 support via ggml-org/llama.cpp#21309, so this resolves the |
* fix(studio): revert llama.cpp default tag to latest The latest ggml-org/llama.cpp release (b8637) now includes Gemma 4 support. Revert the temporary "b8637" pin from #4796 to "latest" so the prebuilt resolver always picks the newest release automatically without needing manual tag bumps. * docs: add comment explaining latest vs master for llama.cpp tag Document in all three files why "latest" is preferred over "master" and when "master" should be used as a temporary override. --------- Co-authored-by: Daniel Han <[email protected]>
1 participant
Summary
b8637is now released with Gemma 4 support (model: support gemma 4 (vision + moe, no audio) ggml-org/llama.cpp#21309)"master"default from fix(studio): build llama.cpp from master for Gemma 4 support #4790 back to a pinned release tagChanges
setup.sh:_DEFAULT_LLAMA_TAG="master"->"b8637"setup.ps1:$DefaultLlamaTag = "master"->"b8637"install_llama_prebuilt.py:DEFAULT_LLAMA_TAGfallback"master"->"b8637"Test plan
UNSLOTH_LLAMA_TAGenv var override still worksunsloth studio updatepicks up the new tag and downloads the prebuilt